GeoFormer: A Geometric Representation Transformer for Change Detection

Deep representation learning has improved automatic remote sensing change detection (RSCD) in recent years. Existing methods emphasize primarily convolutional neural networks (CNNs) or transformer-based networks. However, most of them neither effectively combine CNNs and transformers nor use prior geometric information to refine regions. In this article, a novel geometric representation transformer (GeoFormer) is proposed for high-resolution RSCD. GeoFormer utilizes convolutional information to guide the transformer by employing geometric prior knowledge. Specifically, the proposed GeoFormer consists of three carefully designed components: the geometric-based Swin transformer (Geo-Swin Transformer) encoder, the Laplace attention fusion (LAFusion) module, and the UNet++CD decoder. First, the Geo-Swin Transformer is a novel-designed nonlocal Siamese encoder that combines geometric convolution with a transformer to provide local geometric representation information for remote contextual features. Then, an LAFusion module is proposed to achieve robust bi-temporal feature fusion, which is founded on attention mechanism and edge information. Finally, UNet++CD decodes fine-grained information from the fused features by a dense multiscale upsampling process. Experimental results demonstrate that the proposed GeoFormer performs better than benchmark methods on four change detection datasets (LEVIR-CD, WHU-CD, DSIFN-CD, and CDD) and is able to detect the edges of change regions more precisely.